Syllabus

Introduction: Representing text, Sounds and Images text, speech, image, and video. Signal processing for feature extraction: for Text (BoW), Speech (LPC, Mel-frequency Capstral coefficients, STFT and Wavelet features), Images (HoG, BoVW, FV), Videos (BoVW). Machine Learning basics - Introduction to pattern recognition, Bayesian decision theory, supervised learning from data, parametric and non parametric estimation of density functions, Bayes and nearest neighbor classifiers, introduction to statistical learning theory, empirical risk minimization, discriminant functions, learning linear discriminant functions, Perceptron, linear least squares  regression, LMS algorithm, Supervised and Unsupervised learning, Classification and Regression (linear models), Evaluation metrics, Probability Models and Expectation-Maximization Algorithm, Gaussian Mixture Models, Neural Networks and Deep Learning, Multi-class classification and Multilabel classification, Different kinds of non-linearities, objective functions, and learning methods, ML for Audio Classification, Time Series Analysis, LSTMs, and CNNs, ML for Speech Recognition, Hidden Markov Models, Finite State Transducers and Dynamic Programming, ML for Music Information Retrieval, Latent Variable Models, Matrix Factorization and Signal Separation, ML for Image Processing, Transfer Learning, Attention models, Attribute-based learning, ML for Communication, Deep learning for wireless applications